A First-Principles Framework Based on Virtual Crystal Approximation for Calculating the Lattice Thermal Conductivity of Doped Si1−𝑥Ge𝑥 Alloys

Thermoelectric materials convert waste heat directly into electricity, and their efficiency is strongly influenced by low lattice thermal conductivity. Doping is a widely used strategy to reduce lattice thermal conductivity and enhance thermoelectric performance. However, accurately predicting the thermoelectric properties of doped materials using first-principles calculations is challenging due to the need for large supercells to model the disorder introduced by doping. The Virtual Crystal Approximation (VCA) provides a computationally efficient alternative by treating the doped system as an effective medium with averaged properties, thus avoiding the need for large supercells [1]. Here, we present a first-principles framework based on VCA to calculate the lattice thermal conductivity of doped Si1−xGex alloys. This framework combines Density Functional Theory (DFT) with phonon transport calculations using Phonopy [2, 3] and Phono3py [2, 4] packages, allowing for accurate predictions of lattice thermal conductivity in doped systems. We validate our framework by comparing the calculated lattice thermal conductivity of Si1− x Gex alloys with experimental data, demonstrating good agreement across a range of doping concentrations. Our results show that the VCA-based approach effectively captures the impact of doping on phonon scattering and thermal transport properties. This framework provides an effective approach for improving and design high-performance thermoelectric materials through doping strategies.

Work In Progress